Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Modelling and Forecasting of Electrical and Thermal Energy Units for a MPC based Home Energy Management System

Jacobsson, Måns (2025)
Department of Automatic Control
Abstract
With the shift towards renewable energy, the need for effective energy management grows. Renewable sources introduce variability and unpredictability, leading to volatile energy prices that expose consumers to fluctuations. To mitigate costs, energy storage and forecast models for electricity and thermal energy can be used to make optimal decisions on when to purchase energy from the grid.
This thesis investigates the integration of two forecast models—an energy spot price forecast and an energy consumption forecast—into a Model Predictive Control (MPC) based Home Energy Management System (HEMS) to minimize net energy costs. These models are evaluated against both perfect foresight models as an upper target and baseline persistence... (More)
With the shift towards renewable energy, the need for effective energy management grows. Renewable sources introduce variability and unpredictability, leading to volatile energy prices that expose consumers to fluctuations. To mitigate costs, energy storage and forecast models for electricity and thermal energy can be used to make optimal decisions on when to purchase energy from the grid.
This thesis investigates the integration of two forecast models—an energy spot price forecast and an energy consumption forecast—into a Model Predictive Control (MPC) based Home Energy Management System (HEMS) to minimize net energy costs. These models are evaluated against both perfect foresight models as an upper target and baseline persistence models as a lower benchmark. The models are assessed individually using mean absolute error (MAE) and as part of the MPC to analyze cost reduction.
Results showed that while both the spot price and consumption forecast models performed better than the persistence models in terms of MAE, only the spot price forecast model contributed to reducing net energy costs when integrated into the MPC based HEMS. The consumption forecast model struggled to capture the amplitude and volatility of the data, presumably caused by the included heating consumption. It was concluded that forecasting domestic energy consumption and heating consumption separately (if possible) could improve accuracy and performance. (Less)
Please use this url to cite or link to this publication:
author
Jacobsson, Måns
supervisor
organization
year
type
M2 - Bachelor Degree
subject
report number
TFRT-7670
other publication id
0280-5316
language
English
id
9183996
date added to LUP
2025-02-04 14:00:42
date last changed
2025-02-04 14:00:42
@misc{9183996,
  abstract     = {{With the shift towards renewable energy, the need for effective energy management grows. Renewable sources introduce variability and unpredictability, leading to volatile energy prices that expose consumers to fluctuations. To mitigate costs, energy storage and forecast models for electricity and thermal energy can be used to make optimal decisions on when to purchase energy from the grid.
 This thesis investigates the integration of two forecast models—an energy spot price forecast and an energy consumption forecast—into a Model Predictive Control (MPC) based Home Energy Management System (HEMS) to minimize net energy costs. These models are evaluated against both perfect foresight models as an upper target and baseline persistence models as a lower benchmark. The models are assessed individually using mean absolute error (MAE) and as part of the MPC to analyze cost reduction.
 Results showed that while both the spot price and consumption forecast models performed better than the persistence models in terms of MAE, only the spot price forecast model contributed to reducing net energy costs when integrated into the MPC based HEMS. The consumption forecast model struggled to capture the amplitude and volatility of the data, presumably caused by the included heating consumption. It was concluded that forecasting domestic energy consumption and heating consumption separately (if possible) could improve accuracy and performance.}},
  author       = {{Jacobsson, Måns}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Modelling and Forecasting of Electrical and Thermal Energy Units for a MPC based Home Energy Management System}},
  year         = {{2025}},
}